|
Pannanugget Consulting
Developing and Promoting Chemotypic Methodologies
|
|
|
Background
The basic notion of a chemotype is quite simple. You specify an atom or bond
property, delete those atoms or bonds not having that property and the maximal
fragments that remain are the chemotypes. The concept becomes more general,
and, consequently, more complex, by reducing or enhancing chemotypic resolution.
Chemotypic resolution is reduced by ignoring information in the atom labels
and bond types to obtain, for example, the Bemis and Murcko frameworks or by
collapsing spacer atoms to obtain various reduced-graph chemotypes. It is
enhanced by taking aspects of positioning into account.
The 2003 GRC poster Specifying and Using Hierarchically-Organized Structural Categories, now somewhat dated as to terminology, briefly overviews chemotypes and the way this resolution comes into play in a preliminary comparison of cyclic-system, ring system, side chain and functional group chemotypes as descriptors related to the therapeutic categories of a random selection of 200 structures from the Merck Index. Meqi is first and foremost a research tool. It is designed to quickly explore and provide insight into ways of solving a wide variety of tasks through the use of chemotypes. It is essentially a modifiable, template-based programming language together with a command line facility for viewing and summarizing the output and for reformatting that output in a variety of ways to suit standard data analysis software. Details of the language, the various chemotypic categories Meqi can compute and the notational differentiation of these categories are given in the technical report The Meqi II Programming Template and Its Underlying Chemotypic Concepts. Tutorials
As a template-base programming language, Meqi provides many fill-in-the-slot options. Although any one option is simple, the combinatorial possibilities are quite varied and can be surprising and sometimes perplexing in their implications. In a somewhat analogous way, many computer-assisted drug-design issues are addressed via a number of conceptually simple tasks, many of which can, and occasionally must, be addressed as a chemotypic task. Our goal in constructing these tutorials is to show how a number of these chemotypic tasks can be implemented using only a small subset of the template options. Those tutorials that cover the simplest chemotypic tasks and illustrate the most basic template options will be the first to be made available. However, if you have a particular chemotypic task you would like to see addressed, or are a Meqi user and would care to contribute a tutorial, please feel free to contact us. Finding Isomers and Duplicate Structures Using Uninformative Chemotypic Identifiers. -January, 2009 Characterizing and Contrasting Compound Libraries with Respect to Their Ring Systems and Other Chemotypic Categories. - April, 2009
|